Medication Adherence, Mobile Application
Conditions
Brief summary
Adverse drug events (ADEs) are a leading cause of death in North America, with over 2 million ADEs causing 100,000 deaths every year and millions in economic costs. The majority of ADEs that occur within 30 days of discharge from hospital are related to prescription medications and of these, more than 50% are preventable. Among the causes of ADEs are incomplete and unclear medication lists, and a lack of patient understanding and adherence. Patients are usually discharged from hospital on substantially difference medication regiments than those prior to admission. In the absence of supportive mechanisms that ensure patients' understanding of their medication list, adherence to prescribed medication changes is likely to be reduced and the subsequent risk of ADEs increased. Secondary analyses from a recently conducted randomized controlled trial (RCT) by our team suggest that more than 50% of study patient were non-adherent to at least one in-hospital medication change, and that this non-adherence significantly increased the risk of hospital re-admission and emergency department visits in the 30 days post-discharge. Furthermore, analyses of interview data suggest that non-adherence may be driven by unclear communication with patients about medication changes and the reasons for these changes, as well as difficulties in managing complex dosing schedules and drug regimen information. There is therefore a clear need to implement and evaluate patient support mechanisms that reduce non-adherence to essential changes in therapy following hospitalization. The objective of this project is to conduct a pilot RCT that will evaluate the usability of a medication management mobile application and its efficacy in reducing non-adherence to in-hospital medication changes following discharge. We will randomize 100 patients from the internal medicine unit of the McGill University Health Centre (MUHC) Glen site to either the intervention or control arm. Patients in the control arm will receive usual care (i.e. no medication management support), whereas those in the intervention arm will receive a tablet with the installed mobile application. The application will integrate prescription claims data from the Régie de l'assurance maladie du Québec (RAMQ) for the 3-month period prior to hospital admission with the patient's discharge prescription to generate a patient-friendly medication list, along with details of in-hospital medication changes. The app will also offer a number of features designed to maximize patient understanding and adherence, including pill images, patient-friendly drug monographs, weekly dosing schedules, drug alerts, home refill services, and features that connect with the patient's caregiver and hospital pharmacist. At 1 week post-discharge, study coordinators will conduct a usability assessment to obtain patient feedback on the app (via a technology acceptance questionnaire) and to document usability using the think aloud protocol, which is based on observing and recording patients as they use the application and verbalize any thoughts that might occur to them. Qualitative analysis of recorded and transcribed sessions will then be used to assess the technology's ease of use, user-friendliness, efficiency, and any features that may cause confusion, frustration, or user errors. Non-adherence to in-hospital treatment changes will be assessed by comparing patients' discharge prescriptions with medications dispensed in the 30 days following hospital discharge (obtained from RAMQ pharmacy claims). We will compare, between intervention and control groups, the average number of in-hospital medication changes not adhered to in the 30-day follow-up period. This project will assess a technological intervention that has the potential to improve patient adherence to in-hospital medication changes and may subsequently reduce the occurrence of ADEs. Given the high costs associated with ADEs, this small investment has the potential to incur significant cost savings for the Quebec healthcare system.
Interventions
SAM generates a patient-friendly list of prescribed and dispensed medications and offers several features: * Drug information: Patient-friendly monographs that describe treatment indications and the harms/benefits of medications. * Adherence alerts: Alerts users to non-adherence, such as when they fill prescriptions that were stopped at hospital discharge, don't fill prescribed medications, or buy medications at the incorrect dose. * Side-effect checker: Provides a list of side effects associated with individual medications and the patient's overall drug profile. * Rate my med: Allows patients to rate medications and describe experiences with a medication. Patients can view ratings given by users. * Caregiver connect: Provides patients' caregivers with access to the app. * Pharmacy connect: Allows patients to connect with their hospital pharmacist to ask questions regarding their medications and adherence alerts. * Other: pill images, daily pill reminders, weekly dosing schedule
Sponsors
Study design
Masking description
Due to the nature of the intervention, study participants are not blinded to treatment arm allocation, but data analysts are.
Eligibility
Inclusion criteria
* Discharged home from internal medicine units of the McGill University Health Center * Have a mobile or tablet device with internet connection * Have prescription drug insurance from the provincial health insurer * Prescribed at least one medication at discharge from hospital
Exclusion criteria
* Discharged to a rehabilitation center * Transferred to a non-study unit' * Prognosis of less than 3 months
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Usability of the SAM App | At baseline, prior to hospital discharge | Users were observed as they used the app, following the think aloud protocol. This involved users verbalizing their thoughts as they perform a standardized set of tasks, while being audio-recorded.A field assistant familiar with the SAM app observed and recorded users as they accessed SAM and completed the following standardized tasks: log in, peruse drug information, send the pharmacist a message, add a medication review, use the interaction checker, and use the side effect checker. User performance was observed and registered by a trained research assistant to establish if the user was able to complete the task and to determine the ease/difficulty with which the task was completed, rated in 1-to-4 scale as (1) user can do it perfectly, without any help, (2) user can do it with difficulty but without help, (3) user can do it with difficulty and help, and (4) user cannot do it. |
| User Acceptance of and Satisfaction With the SAM App | 2 weeks post-discharge | Participants in the intervention arm completed the Technology Acceptance Model (TAM) Questionnaire 2 weeks following discharge from hospital. The questionnaire measures users' perceived ease of use and perceived usefulness of the SAM app and their intention to use the app. The TAM questionnaire is self-administered, and consists of 38 questions on a Likert scale of 1 to 7, with responses being strongly agree (7), moderately agree (6), somewhat agree (4), neutral (4), somewhat disagree (3), moderately disagree (2), and strongly disagree (1). Responses higher on the Likert scale typically reflected higher acceptance, usability, perceived usefulenss of the app, and intent to use the app from users' perspectives. |
| Proportion of Medication Changes That Are Adhered to | 30 days post-discharge | Pharmacy claims data from the public insurance provider will be used to identify community medications dispensed to participating patients in the 30 days following hospital discharge. These data will be compared with the patient's hospital discharge prescription to identify non-adherence to in-hospital medication changes. Non-adherence to in-hospital treatment changes in the 30-day post-discharge period will be defined as (1) failure to fill newly prescribed medications, (2) re-filling a previous prescription that was stopped at discharge, or (3) filling a previous prescription whose dosage was modified at discharge, but was dispensed post-discharge at the incorrect daily dose. For each type of medication change, the number of changes among all patients will be counted (e.g. number of newly prescribed medications among the patient population), and the number of those that were properly adhered to will then be used to calculate the proportion of medication changes that were adhered to. |
Countries
Canada
Participant flow
Participants by arm
| Arm | Count |
|---|---|
| SAM Intervention Group At discharge from hospital, patients (and/or caregivers) received access to Smart About Meds (SAM), a medication management mobile application that has been developed by the McGill Clinical and Health Informatics (MCHI) Research Group. | 23 |
| Usual Care Group At discharge from hospital, patients were provided with a written discharge prescription to be filled at their community pharmacy, and may or may not have received written or verbal instructions about changes made to therapy. | 26 |
| Total | 49 |
Baseline characteristics
| Characteristic | Usual Care Group | Total | SAM Intervention Group |
|---|---|---|---|
| Age, Continuous | 68.6 years | 64.6 years | 54.6 years |
| ED visit in 3 months prior to admission | 14 Participants | 29 Participants | 15 Participants |
| Hospitalization in 3 months prior to admission | 5 Participants | 10 Participants | 5 Participants |
| Number of discontinued medications at discharge | 2.0 medications | 2.0 medications | 2.0 medications |
| Number of dose changes at discharge | 1.0 medications | 1.0 medications | 1.0 medications |
| Number of medications changed at discharge | 7.0 medications | 7.0 medications | 6.0 medications |
| Number of medications continued at discharge | 11.5 medications | 9.0 medications | 6.0 medications |
| Number of medications prescribed at discharge | 15.5 medications | 15.0 medications | 13.0 medications |
| Number of new medications at discharge | 3.0 medications | 3.0 medications | 3.0 medications |
| Participant has caregiver | 7 Participants | 8 Participants | 1 Participants |
| Race and Ethnicity Not Collected | — | 0 Participants | — |
| Sex: Female, Male Female | 8 Participants | 19 Participants | 11 Participants |
| Sex: Female, Male Male | 18 Participants | 30 Participants | 12 Participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk |
|---|---|---|
| deaths Total, all-cause mortality | 0 / 0 | 0 / 0 |
| other Total, other adverse events | 0 / 0 | 0 / 0 |
| serious Total, serious adverse events | 0 / 0 | 0 / 0 |
Outcome results
Proportion of Medication Changes That Are Adhered to
Pharmacy claims data from the public insurance provider will be used to identify community medications dispensed to participating patients in the 30 days following hospital discharge. These data will be compared with the patient's hospital discharge prescription to identify non-adherence to in-hospital medication changes. Non-adherence to in-hospital treatment changes in the 30-day post-discharge period will be defined as (1) failure to fill newly prescribed medications, (2) re-filling a previous prescription that was stopped at discharge, or (3) filling a previous prescription whose dosage was modified at discharge, but was dispensed post-discharge at the incorrect daily dose. For each type of medication change, the number of changes among all patients will be counted (e.g. number of newly prescribed medications among the patient population), and the number of those that were properly adhered to will then be used to calculate the proportion of medication changes that were adhered to.
Time frame: 30 days post-discharge
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| SAM Intervention Group | Proportion of Medication Changes That Are Adhered to | Proportion of all medication changes adhered to | 0.837 proportion of medication changes |
| SAM Intervention Group | Proportion of Medication Changes That Are Adhered to | Proportion of all dose changes adhered to | 0.909 proportion of medication changes |
| SAM Intervention Group | Proportion of Medication Changes That Are Adhered to | Proportion of all new medications adhered to | 0.727 proportion of medication changes |
| SAM Intervention Group | Proportion of Medication Changes That Are Adhered to | Proportion of all discontinued medications adhered to | 0.935 proportion of medication changes |
| Usual Care Group | Proportion of Medication Changes That Are Adhered to | Proportion of all discontinued medications adhered to | 0.943 proportion of medication changes |
| Usual Care Group | Proportion of Medication Changes That Are Adhered to | Proportion of all medication changes adhered to | 0.778 proportion of medication changes |
| Usual Care Group | Proportion of Medication Changes That Are Adhered to | Proportion of all new medications adhered to | 0.617 proportion of medication changes |
| Usual Care Group | Proportion of Medication Changes That Are Adhered to | Proportion of all dose changes adhered to | 0.818 proportion of medication changes |
Usability of the SAM App
Users were observed as they used the app, following the think aloud protocol. This involved users verbalizing their thoughts as they perform a standardized set of tasks, while being audio-recorded.A field assistant familiar with the SAM app observed and recorded users as they accessed SAM and completed the following standardized tasks: log in, peruse drug information, send the pharmacist a message, add a medication review, use the interaction checker, and use the side effect checker. User performance was observed and registered by a trained research assistant to establish if the user was able to complete the task and to determine the ease/difficulty with which the task was completed, rated in 1-to-4 scale as (1) user can do it perfectly, without any help, (2) user can do it with difficulty but without help, (3) user can do it with difficulty and help, and (4) user cannot do it.
Time frame: At baseline, prior to hospital discharge
Population: This analysis population included 31 patients who consented to participate in the pilot and who were randomized to the intervention group and baseline and were successfully administered the questionnaire. It includes some individuals who were withdrawn from the study following randomization for various reasons (e.g. withdrew consent, discharged to a rehabilitation center, transferred to a non-study unit) and not included in other outcome analysis populations.
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| SAM Intervention Group | Usability of the SAM App | Severity scale mean, login | 1.14 score on a scale | Standard Deviation 0.51 |
| SAM Intervention Group | Usability of the SAM App | Severity scale mean, view drug information | 1.32 score on a scale | Standard Deviation 0.74 |
| SAM Intervention Group | Usability of the SAM App | Severity scale mean, send a message | 1.35 score on a scale | Standard Deviation 0.65 |
| SAM Intervention Group | Usability of the SAM App | Severity scale mean, review a medication | 1.81 score on a scale | Standard Deviation 1.09 |
| SAM Intervention Group | Usability of the SAM App | Severity scale mean, check side effects of a specific medication | 1.13 score on a scale | Standard Deviation 0.45 |
| SAM Intervention Group | Usability of the SAM App | Severity scale mean, check side effect profile of all medications (side effect checker) | 1.71 score on a scale | Standard Deviation 0.88 |
| SAM Intervention Group | Usability of the SAM App | Severity scale mean, use interaction checker | 1.17 score on a scale | Standard Deviation 0.37 |
User Acceptance of and Satisfaction With the SAM App
Participants in the intervention arm completed the Technology Acceptance Model (TAM) Questionnaire 2 weeks following discharge from hospital. The questionnaire measures users' perceived ease of use and perceived usefulness of the SAM app and their intention to use the app. The TAM questionnaire is self-administered, and consists of 38 questions on a Likert scale of 1 to 7, with responses being strongly agree (7), moderately agree (6), somewhat agree (4), neutral (4), somewhat disagree (3), moderately disagree (2), and strongly disagree (1). Responses higher on the Likert scale typically reflected higher acceptance, usability, perceived usefulenss of the app, and intent to use the app from users' perspectives.
Time frame: 2 weeks post-discharge
Population: This analysis population includes individuals who consented to participate in the pilot, were randomized to the intervention group at baseline, and who at 2 week post-discharge agreed to complete the TAM questionnaire. Only 12 of the 23 participants in the intervention group were reachable and agreed to complete the questionnaire at this point during follow-up.
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| SAM Intervention Group | User Acceptance of and Satisfaction With the SAM App | TAM domain: behavioral intention | 5.17 score on a scale | Standard Deviation 2.12 |
| SAM Intervention Group | User Acceptance of and Satisfaction With the SAM App | TAM domain: perceived usefulness | 5.29 score on a scale | Standard Deviation 2.11 |
| SAM Intervention Group | User Acceptance of and Satisfaction With the SAM App | TAM domain: subjective norm | 3.81 score on a scale | Standard Deviation 1.74 |
| SAM Intervention Group | User Acceptance of and Satisfaction With the SAM App | TAM domain: relevance | 4.86 score on a scale | Standard Deviation 1.78 |
| SAM Intervention Group | User Acceptance of and Satisfaction With the SAM App | TAM domain: output quality | 5.44 score on a scale | Standard Deviation 1.65 |
| SAM Intervention Group | User Acceptance of and Satisfaction With the SAM App | TAM domain: result demonstrability | 5.06 score on a scale | Standard Deviation 1.82 |
| SAM Intervention Group | User Acceptance of and Satisfaction With the SAM App | TAM domain: perceived ease of use | 5.65 score on a scale | Standard Deviation 2.02 |
| SAM Intervention Group | User Acceptance of and Satisfaction With the SAM App | TAM domain: computer self-efficacy | 5.60 score on a scale | Standard Deviation 2 |
| SAM Intervention Group | User Acceptance of and Satisfaction With the SAM App | TAM domain: perceived enjoyment | 5.34 score on a scale | Standard Deviation 1.97 |